AI companies have spent the last few years insisting AI is here to help. It’ll help us write faster. Help us design quicker. Help us think better.
This month, it became clear that AI is no longer just helping, it’s starting to stand between us and the thing we’re trying to do. It’s positioning itself between consumers and products, between the creators and their audiences.
Which is fine. Every technology does this eventually. The problem starts when we pretend nothing has changed.
So, who’s buying?
Not so long ago, the digital economy was built around interfaces. Websites, apps, feeds and funnels were carefully designed paths where brands shepherded people from curiosity to conversion like sheepdogs. We got damn good at it, too.
Now along comes the Universal Commerce Protocol, quietly suggesting that the future of shopping doesn’t actually involve browsing at all. Instead of us falling in love with a product that gets interjected in the middle of our evening doomscroll, AI agents will discover, compare, decide, and check out on behalf of us — inside conversations, not websites.
In other words: as a brand, your next “customer” might not have eyes. And you might want to skip those emotional triggers, too.
This doesn’t kill branding, despite what some LinkedIn prophets might tell you. It does, however, mean that a lot of what we’ve spent decades perfecting, visual persuasion, micro-copy heroics, lovingly crafted landing pages, may become invisible and irrelevant to the systems actually making decisions.
Taste is still relevant. Clarity is now existential.
At the same time, the industry keeps flirting with the fantasy that AI can simply do the work for us. Finally, at last, no more messy humans and nagging creatives. Reality, unfortunately, remains stubborn.
When researchers tested AI against real, paid freelance tasks – the kind clients actually approve and pay for – the success rate was a humbling two and a half percent. Not because the outputs were insane, but because they were almost right. Which, as anyone who’s ever presented to a client knows, is the most dangerous category of work.
Inside agencies, the pattern is already familiar. AI happily generates 200 options in seconds. Then a human spends hours deciding which two aren’t brand-damaging, legally risky, or aesthetically offensive. The work didn’t disappear. It just changed clothes. What used to be invisible judgment, perfected taste, is now the job.
Cue the backlash.
As generative content floods feeds, audiences are reacting the way humans always do when something starts to feel mass-produced: with boredom, suspicion, and the occasional mob armed with torches and pitchforks. “AI Slop” has become the unofficial term of the moment. It’s content that technically works, emotionally doesn’t, and will be forgotten before the scroll thumb resets. We have seen mistakes in approved material that would never have been green-lit by a creative team (or a client) it it had been done in-camera.
Music research landed on the same conclusion in a more polite tone. AI-generated songs can sound fine, but listeners consistently find human-created music more engaging. More dynamic. More alive. Apparently, emotional resonance is still not a checkbox you can toggle on.
This is where a mildly unfashionable idea returns: taste. Not as vibes or the frail ego of an art director in a fisherman’s cap, but as a practical, commercial advantage. When content is infinite, choosing what not to make becomes the differentiator. Like Mikko Hovi put it, “these tools give you the chance to do anything you can imagine, but you shouldn’t just because you can”.
AI goes physical
Then there’s physical AI. Robots wandering CES floors. Assistants leaving the screen and entering the room. Brand activations with legs. The recent demos are impressive. The reliability was… aspirational. Physical AI has a unique talent for turning hype into embarrassment at record speed. When software fails, it glitches. When robots fail, they stare into the middle distance like they’ve seen things.
For brands, the opportunity is real — embodied experiences, new forms of interaction, presence beyond the phone. But the margin for error shrinks dramatically once AI has mass, volume, and gravity.
Use it or lose it
Underneath all of this sits a quieter, more uncomfortable theme: cognitive outsourcing. OECD warns about “false mastery” in education, the illusion of understanding created when AI fills gaps too smoothly. This applies just as much to creative and strategic work. When systems complete our thoughts, it becomes harder to notice what we’re no longer doing ourselves.
Experiments like AI-only social networks underline the point in stranger ways. Agents talking to agents, forming belief systems, governing themselves, occasionally leaking credentials all over the internet. Meanwhile, humans watch from the sidelines, half-amused, half-concerned.
Capability is rising fast. Reliability is jogging behind, waving.
So no, AI isn’t killing creativity. It’s doing something more annoying: forcing clarity.
Execution is cheap now. Endless options are a keystroke away. And leadership — creative or otherwise — is shifting from making things to deciding what deserves to exist at all. Which brings us to the practical bit.
Three things you can actually do:
1. Design for agents, not just humans
Your brand needs to make sense to systems that read, summarise, recommend, and decide, without the emotional or cultural layer that we so often rely on. That means clearer product logic, tighter narratives, and fewer “you had to be there” ideas.
2. Invest in taste like it’s infrastructure
If AI gives everyone speed, your advantage is judgment. Hire for it, train it, protect it. Make it visible. This is not a soft skill anymore, it’s a moat. Your brand should have an opinion on matters of taste, so should you. Demand clarity and vision.
3. Be explicit about what stays human
Not everything should be automated just because it can be. Decide where intuition, ethics, and responsibility live – and say it out loud. Audiences notice when you don’t. There is a craving for real things, audiences reward you for making them.